Nvidia Jetson TK1 experiements

Decided to try and port some mining software that uses CUDA and cryptonight crypto (XMR) to the Nvidia Jetson TK1 development board that I have acquired recently to study.

This device has the following specificationsProcessor:

  • GPU consisting of 192 ALUs using Kepler technology
  • 28 nm HPM process
  • Released in Q2 2014
  • Power consumption: 8 watts

I have ported the xmrMiner (https://github.com/xmrMiner) onto this device (since it does not build without modifications to the code on the Jetson TK1)  as it can use only CUDA 6.5 since it is 32bit ARMv7 architecture and Nvidia dropped 32bit support for CUDA since then.

The port consisted of modifying the GPU assembler functions in cryptonight/cuda_cryptonight_core.cu   and some ARM declarations in sha2.c

Required version of cmake was compiled via pkgsrc-current on the Jetson (pkgsrc is awsome here) with no major issues.

I have created a git repo for this specific fork here


Below is the GPU miner running  – benchmarking gave approx 15 H/sec

With a combination of xmrig ( https://github.com/xmrig/xmrig) you can get additional 12 – 15 H/sec from the Cortex ARM CPU, so overall  performance would be around 25 – 30 H/sec from both CUDA GPU and the CPU.

This was a nice introduction into CUDA programming and ARM porting that I have enjoyed very much over this weekend.




About astr0baby

Please run Adblock or similar... we have been told to do so since Carl Sagan wrote the Contact .
This entry was posted in Uncategorized. Bookmark the permalink.

1 Response to Nvidia Jetson TK1 experiements

  1. Pingback: Nvidia Jetson TK1 experiements | 懒得折腾

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.